Datasets:

google-research-datasets
/

nq_open

question stringlengths 19 100	answer sequencelengths 1 25
where did they film hot tub time machine	[ "Fernie Alpine Resort" ]
who has the right of way in international waters	[ "Neither vessel" ]
who does annie work for attack on titan	[ "Marley" ]
when was the immigration reform and control act passed	[ "November 6, 1986" ]
when was puerto rico added to the usa	[ "1950" ]
who has been chosen for best supporting actress in 64 national filmfare award	[ "Zaira Wasim" ]
which side of the white house is the front	[ "North" ]
names of the metropolitan municipalities in south africa	[ "Mangaung Metropolitan Municipality", "Nelson Mandela Bay Metropolitan Municipality", "eThekwini Metropolitan Municipality", "City of Tshwane Metropolitan Municipality", "City of Johannesburg Metropolitan Municipality", "Buffalo City Metropolitan Municipality", "City of Ekurhuleni Metropolitan Municipality" ]
who's hosting the super bowl in 2019	[ "Atlanta, Georgia" ]
in which year vivo launch its first phone in india	[ "2014" ]
where does it talk about mary magdalene in the bible	[ "New Testament" ]
who carries the nuclear football for the president	[ "aide-de-camp" ]
what is the origin of the name cynthia	[ "Greek" ]
who is the guy who voiced disney channel	[ "\"Buzz\" Brainard", "Cameron" ]
what's the legal marriage age in new york	[ "18" ]
when do the red hot chili peppers tour	[ "March 2018" ]
who plays mavis in the movie hotel transylvania	[ "Sadie Sandler", "Selena Gomez" ]
what is the channel number for cartoon network on spectrum	[ "768 (HD)", "68 (SD)" ]
when are the fa cup semi finals played	[ "the new Wembley Stadium" ]
when did the ipod touch 6 gen came out	[ "July 15, 2015" ]
west ham players in the 1966 world cup	[ "Martin Peters", "Geoff Hurst", "Bobby Moore" ]
who sings the theme song for miami vice	[ "Jan Hammer" ]
who sings should have seen it in color	[ "Jamey Johnson" ]
who starred in the movie deep water horizon	[ "Kurt Russell", "Dylan O'Brien", "John Malkovich", "Kate Hudson", "Gina Rodriguez", "Mark Wahlberg" ]
when did justin bieber perform at madison square garden	[ "August 31, 2010" ]
protozoan classification used to be based on their means of	[ "locomotion" ]
who was the first one to migrate to madinah	[ "Islamic prophet Muhammad" ]
ice sheets and tundra are typical of which koppen climate category	[ "polar" ]
what is the size of the angles of an equilateral triangle	[ "60°" ]
when does beaty and the beast take place	[ "Rococo-era" ]
who sings done for me with charlie puth	[ "American singer Kehlani" ]
what counties are in the 6th congressional district in ga	[ "Fulton", "DeKalb", "Cobb" ]
who was heading indian army during bangladesh libration war 1971	[ "Sam Manekshaw" ]
who wrote the music for somewhere in time	[ "John Barry" ]
who sings the marriage of figaro in shawshank redemption	[ "Edith Mathis", "Gundula Janowitz" ]
who did chris long play football for last year	[ "Eagles" ]
when did harry potter and the deathly hallows come out	[ "21 July 2007" ]
how many episode of my hero academia season 2	[ "25" ]
during the rebuttal section of a formal debate which team speaks first	[ "the affirmative team" ]
when was the last amendment made to the us constitution	[ "May 5, 1992" ]
starch based food to be made into plastic	[ "corn starch" ]
when did north korea launch the first missile	[ "1984" ]
what mineral in 4 ga sandstones was dated to determine the age of the earth	[ "zircon" ]
when is the last time a total eclipse happened in the us	[ "August 21, 2017" ]
when did india participate in olympics for first time	[ "1900" ]
when is it revealed that luke and leia are siblings	[ "In Return of the Jedi" ]
what is the population of farmington new mexico	[ "41,629" ]
what was the last year they made the gmc envoy	[ "2008" ]
new amsterdam was the main settlement in the dutch colony of	[ "New Netherland" ]
in the bible who was thrown into the fiery furnace	[ "Meshach", "Abednego", "Shadrach" ]
how does the oral cavity help the digestive system	[ "produces a lubricating mucus" ]
who was alf married to in home and away	[ "Ailsa Stewart" ]
when did miss usa last win miss universe	[ "2012" ]
when did the shooting happen in greys anatomy	[ "May 20, 2010" ]
who sings in there's something about mary	[ "Jonathan Richman" ]
when will new tappan zee bridge be completed	[ "the end of 2018" ]
where is the rio grande river located in jamaica	[ "parish of Portland" ]
where did the us get the statue of liberty	[ "the people of France" ]
where is the new tottenham stadium being built	[ "the London Borough of Haringey" ]
who won icc t 20 world cup 2016	[ "West Indies" ]
who does jessica marry in the merchant of venice	[ "Lorenzo" ]
when were the triplets born on this is us	[ "1980" ]
what degree do you get out of medical school	[ "Doctor of Medicine (MD" ]
what was the source of alpha particles in rutherford's experiment	[ "helium nuclei" ]
who plays glenn on the good doctor episode 7	[ "Titus Welliver" ]
who does patrick stewart play in star trek	[ "Captain Jean-Luc Picard" ]
real name of ginny weasley in harry potter	[ "Bonnie Francesca Wright" ]
what animal is mr. big in the movie zootopia	[ "Arctic shrew" ]
who coined the phrase what happens in vegas	[ "R&R Partners", "ad agency" ]
when did the suite life of zack and cody first air	[ "March 18, 2005" ]
who began the mass printing of bibles five centuries ago	[ "Johannes Gutenberg" ]
who is next in line if president dies	[ "the vice president" ]
what was the famous window called in the notre dame cathedral in paris france	[ "rose window" ]
what is reagan's job in new girl	[ "pharmaceutical sales rep" ]
who played raoul in the phantom of the opera	[ "Patrick Wilson" ]
what are the two main political parties in france	[ "The Republicans", "the Socialist Party" ]
when did the golden knights join the nhl	[ "2017" ]
how many episodes in season 1 of the oc	[ "27" ]
where are the upper motor neurons that control skeletal muscles found	[ "the precentral gyrus" ]
what type of government structure does the us have	[ "constitutional republic" ]
when did apple computer change to apple inc	[ "January 9, 2007" ]
who won the mens money in the bank match	[ "Braun Strowman" ]
the depth of the wave base is approximately	[ "half the wavelength" ]
cities in washington state that start with o	[ "Ocean Shores", "Oak Harbor", "Orting", "Othello", "Olympia", "Oroville", "Okanogan", "Omak", "Oakville" ]
when is tornado season in the united states	[ "March through June" ]
when did the olympics start every 2 years	[ "1994" ]
who says the opening of law and order svu	[ "Steven M. Zirnkilton" ]
who has highest number of subscribers on youtube	[ "PewDiePie" ]
when did they launch the international space station	[ "1998" ]
who played willy wonka in the original movie	[ "Gene Wilder" ]
who wrote the song i drive your truck	[ "Jimmy Yeary", "Jessi Alexander", "Connie Harrington" ]
where did the idea of a messiah come from	[ "Judaism" ]
who was the film chariots of fire about	[ "Eric Liddell", "Harold Abrahams" ]
membership in the european union requires countries to have which type of government	[ "democracy" ]
when did the life of pablo come out	[ "February 14, 2016" ]
who played freddy krueger in the 2010 nightmare on elm street	[ "Jackie Earle Haley" ]
who sings dont you wish your girlfriend was hot like me	[ "The Pussycat Dolls" ]
who did lebron james play for before the cleveland cavaliers	[ "Miami Heat" ]
where does the reed go on a saxophone	[ "the mouthpiece" ]
who stars in the movie my cousin rachel	[ "Rachel Weisz", "Holliday Grainger", "Pierfrancesco Favino", "Iain Glen", "Sam Claflin" ]

End of preview. Expand in Dataset Viewer.

Dataset Card for nq_open

Dataset Summary

The NQ-Open task, introduced by Lee et.al. 2019, is an open domain question answering benchmark that is derived from Natural Questions. The goal is to predict an English answer string for an input English question. All questions can be answered using the contents of English Wikipedia.

Supported Tasks and Leaderboards

Open Domain Question-Answering, EfficientQA Leaderboard: https://ai.google.com/research/NaturalQuestions/efficientqa

Languages

English (en)

Dataset Structure

Data Instances

{
    "question": "names of the metropolitan municipalities in south africa",
    "answer": [
        "Mangaung Metropolitan Municipality",
        "Nelson Mandela Bay Metropolitan Municipality",
        "eThekwini Metropolitan Municipality",
        "City of Tshwane Metropolitan Municipality",
        "City of Johannesburg Metropolitan Municipality",
        "Buffalo City Metropolitan Municipality",
        "City of Ekurhuleni Metropolitan Municipality"
    ]
}

Data Fields

question - Input open domain question.
answer - List of possible answers to the question

Data Splits

Train : 87925
validation : 3610

Dataset Creation

Curation Rationale

[Needs More Information]

Source Data

Initial Data Collection and Normalization

Natural Questions contains question from aggregated queries to Google Search (Kwiatkowski et al., 2019). To gather an open version of this dataset, we only keep questions with short answers and discard the given evidence document. Answers with many tokens often resemble extractive snippets rather than canonical answers, so we discard answers with more than 5 tokens.

Who are the source language producers?

[Needs More Information]

Annotations

Annotation process

[Needs More Information]

Who are the annotators?

[Needs More Information]

Personal and Sensitive Information

[Needs More Information]

Considerations for Using the Data

Social Impact of Dataset

[Needs More Information]

Discussion of Biases

Evaluating on this diverse set of question-answer pairs is crucial, because all existing datasets have inherent biases that are problematic for open domain QA systems with learned retrieval. In the Natural Questions dataset the question askers do not already know the answer. This accurately reflects a distribution of genuine information-seeking questions. However, annotators must separately find correct answers, which requires assistance from automatic tools and can introduce a moderate bias towards results from the tool.

Other Known Limitations

[Needs More Information]

Additional Information

Dataset Curators

[Needs More Information]

Licensing Information

All of the Natural Questions data is released under the CC BY-SA 3.0 license.

Citation Information

@article{doi:10.1162/tacl\_a\_00276,
    author = {Kwiatkowski, Tom and Palomaki, Jennimaria and Redfield, Olivia and Collins, Michael and Parikh, Ankur and Alberti, Chris and Epstein, Danielle and Polosukhin, Illia and Devlin, Jacob and Lee, Kenton and Toutanova, Kristina and Jones, Llion and Kelcey, Matthew and Chang, Ming-Wei and Dai, Andrew                         M. and Uszkoreit, Jakob and Le, Quoc and Petrov, Slav},
    title = {Natural Questions: A Benchmark for Question Answering Research},
    journal = {Transactions of the Association for Computational Linguistics},
    volume = {7},
    number = {},
    pages = {453-466},
    year = {2019},
    doi = {10.1162/tacl\_a\_00276},
    URL = { 
            https://doi.org/10.1162/tacl_a_00276
        },
    eprint = { 
            https://doi.org/10.1162/tacl_a_00276
        
        },
    abstract = { We present the Natural Questions corpus, a question answering data set. Questions consist of real anonymized, aggregated queries issued to the Google search engine. An annotator is presented with a question along with a Wikipedia page from the top 5 search results, and annotates a long answer (typically a paragraph) and a short answer (one or more entities) if present on the page, or marks null if no long/short answer is present. The public release consists of 307,373 training examples with single annotations; 7,830 examples with 5-way annotations for development data; and a further 7,842 examples with 5-way annotated sequestered as test data. We present experiments validating quality of the data. We also describe analysis of 25-way annotations on 302 examples, giving insights into human variability on the annotation task. We introduce robust metrics for the purposes of evaluating question answering systems; demonstrate high human upper bounds on these metrics; and establish baseline results using competitive methods drawn from related literature. }
}

@inproceedings{lee-etal-2019-latent,
    title = "Latent Retrieval for Weakly Supervised Open Domain Question Answering",
    author = "Lee, Kenton  and
      Chang, Ming-Wei  and
      Toutanova, Kristina",
    booktitle = "Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics",
    month = jul,
    year = "2019",
    address = "Florence, Italy",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/P19-1612",
    doi = "10.18653/v1/P19-1612",
    pages = "6086--6096",
    abstract = "Recent work on open domain question answering (QA) assumes strong supervision of the supporting evidence and/or assumes a blackbox information retrieval (IR) system to retrieve evidence candidates. We argue that both are suboptimal, since gold evidence is not always available, and QA is fundamentally different from IR. We show for the first time that it is possible to jointly learn the retriever and reader from question-answer string pairs and without any IR system. In this setting, evidence retrieval from all of Wikipedia is treated as a latent variable. Since this is impractical to learn from scratch, we pre-train the retriever with an Inverse Cloze Task. We evaluate on open versions of five QA datasets. On datasets where the questioner already knows the answer, a traditional IR system such as BM25 is sufficient. On datasets where a user is genuinely seeking an answer, we show that learned retrieval is crucial, outperforming BM25 by up to 19 points in exact match.",
}